How Human Insight Enhances AI-Driven Marketing Personalization
- Read original article here
When it comes to personalization in marketing, AI is a great tool but it is not a replacement for human intuition, insightfulness or empathy.
“A hammer or saw doesn't replace the carpenter,” said Jonathan Nelson, senior digital marketing manager at the American Marketing Association.
“It's lacking that intuition. That's where I think the human element is really important for novel thought. AI is really good at creating a collage, but humans have the ability to create something entirely new and original.”
If the goal was to convince another machine it needs an update, then AI would be perfect. But humans are not machines. And since the intent of marketing is to influence how other humans think, feel and act (often based on a set of input and influences that are not sequential or necessarily logical), it still requires another human to understand the emotional impact a marketing campaign or offer will have, said Kelsey Robinson, a senior partner at McKinsey.
“There's an art to how we actually influence customers,” she said. “And it's not always about selling the best product right now. There's parts of personalization and marketing that are just not going to be in a data feedback loop.”
Apple is a good example of this dynamic in action. Over the past decade, many companies have adopted full-funnel marketing, which is a purely data-driven approach to reaching customers. And while many succeed in presenting the next-best-offer on a consistent basis, what’s often neglected is the front of the funnel: generating demand through brand building. Apple, on the other hand, has never neglected its brand. The brand typifies high-quality products that put customers first. This loyalty building helped create the world's first trillion dollar company in 2018 and first $3 trillion company in early 2022.
Related Article: Why Marketing Leaders Need to Harness the Power of AI Now
What the current AI tools are good for is understanding what people buy, when they buy it, how they like to be reached (online, mobile, text, etc.) and at what time of day. At the moment, AI is not particularly useful in helping companies contextualize their offers. This requires a level of scalable content creation that has just been too difficult to achieve in the past.
“That's one of the areas that we've been talking to most clients about,” said Robinson. “If you really want to do personalization at scale ... you actually need a ton of variation in how you communicate with them.”
As they get ever-more sophisticated, easy to use and capable, generative AI tools should make it easier than ever to develop a broad range of personalized content including digital ads, social media posts, email campaigns and the art to support them, she said. And, with ChatGPT’s recent announcement that it is now allowing organizations to leverage its capabilities via plugins, this capability is one step closer to reality.
“With Gen-AI [generative AI] I could do a couple of things,” Robinson said. “I could vary the way I present [products] to you in a nuanced way. If I know what kind of tennis player you are, I could ... say, ‘Okay, for a starting tennis player to get off to a great start, here's three product recommendations for you.’ And then I could have a description on each one: Here's what reviewers have said about this product and why. It's almost like you're replicating what a great in-store associate would do. With Gen-AI, I could do that at scale.”
The ability to generate so much content and iterate on different ideas should prove to be beneficial to creatives, as well, said Robinson. Spinning ideas endlessly to come up with new variations is extremely tiring and something that creatives aren't necessarily all that excited to do. Allowing generative AI to do the heavy lifting in this process, should allow creatives the space they need to actually be more creative and do what humans are particularly good at: coming up with things that are innovative and novel.
As well as creating content for specific audiences and individuals, Nelson sees the next generation of AI being particularly helpful in figuring out who your audience is and then segmenting that audience so that the content you do create using generative AI at scale has some place to go.
“Even with the ability to A/B test at scale quickly, it will still come down to human decision-makers to decide which ads are presented to which constituencies. “The element of feeling, emotion is the biggest thing missing [from AI],” said Robinson. “In every process, there's a decision where my gut tells me, ‘These are the best choices’.”
